Full opinion text
MEMORANDUM ORDER PATRICK E. HIGGINBOTHAM, District Judge. The special master has filed a report recommending back pay awards, including interest to July 1, 1980, totaling $208,148. The method by which these awards were computed is complex, and will be discussed in some detail in the following section. I. The Master’s Report. The master first computed a quantity designated the “weekly differential” for each year for the positions of assistant manager, manager, and produce manager. This number consisted of the difference between the median salary for males in the respective managerial positions and an estimate of the average salary earned by the claimants. The average claimant salary was used rather than the actual salary for each claimant because the master believed that higher-paid claimants should not be penalized for their presumably greater productivity. The weekly differential was then multiplied by the number of weeks each claimant would have served in a managerial position absent discrimination. The latter figure was computed by assuming that each claimant would have been promoted after the median length of time served by male employees before promotion. Where claimants aspired to both the assistant manager/manager positions and the produce manager position, the promotion track resulting in the larger back pay award was used. The promotion ladder was assumed to begin on July 2, 1965, when Title VII became effective, rather than two years before filing of the EEOC charge, as urged by the defendant. This resulted in treatment of most claimants as incumbents in managerial positions from the beginning of the back-pay period. The master did not make any allowance for the possibility that one or more claimants would absent discrimination have reached supervisory rank. While he found it likely that at least one claimant would have been so promoted, he found it impossible, or at least extremely difficult, to determine: (1) which claimants) would have been promoted; (2) to which position(s) they would have been promoted; and (3) the median time before promotion to supervisory rank. He also noted that, with one exception, all male supervisors had (unlike claimants) had managerial experience before coming to Shop Rite. The result of applying hypothetical promotion times to the weekly differential was designated the “Vanilla Back Pay Award.” The respective awards, which total $200,605, are shown in column 1 of the appendix. The master determined that bonuses for employees in the assistant manager and manager positions were isolated occurrences, and recommended that no provision be made for bonuses in such positions. He recommended inclusion of a bonus ranging from $4 to $10 weekly in the produce manager’s salary computations. The master recommended use of a 6% annual interest rate, compounded weekly, for an effective annual rate of 6.18%. He considered and rejected inclusion of an additional amount as an inflation factor, on the ground that Title VII claimants should not fare better in this regard than other judgment creditors. The Vanilla Awards adjusted for interest to date are shown in column 4 of the appendix. The master analyzed a variety of tax effects on the back pay awards, and concluded that no adjustment should be made. He determined that the detriment to the claimants due to lump-sum taxation of the awards in the year of receipt was roughly counterbalanced by the award of compound interest on the amounts which in reality would have been paid to the tax collector in the year in which earned. The master also concluded that the effects of FICA and unemployment taxes should be ignored. The most troublesome problem facing the master was the treatment to be given the fact that male managerial employees were typically terminated (whether voluntarily or involuntarily is not known) after a short tenure in managerial positions. Plaintiffs urged that this fact be ignored altogether, while defendant urged that each claimant’s hypothetical tenure be cut short after the median male tenure had elapsed, even if this meant that some claimants would have been hypothetically terminated before the back-pay period began. The master employed two basic approaches to take this factor into account. First, he computed the rough odds of a particular claimant being terminated after varying times in a given position. These odds were then factored into the week-by-week computations of the Vanilla Awards: the weekly differential was multiplied by the estimated probability that the claimant would have remained in the position during the week in question. To prevent some claimants’ awards from being sharply reduced even at the beginning of the period, the master did not apply a probability factor to the weekly differential until the median number of weeks, shown in n.3, had elapsed from the beginning of the period. The results of these calculations are shown in column 2 of the appendix, and the awards so computed (with interest to July 1, 1980) total $236,073.69. In the master’s second approach, which he ultimately adopted, the aggregate Vanilla Award was first reduced by 40%, then multiplied by an interest factor. This figure ($173,870) was then used as an aggregate award for 11 “major” claimants. Each claimant’s share of this award was determined by multiplying the number of weeks “worked” as manager, plus % of the weeks worked as assistant manager and Vs of the weeks worked as clerk, by the claimant’s average weekly wage as clerk, and dividing by the total of this quantity for all claimants. In addition to the recommended awards computed by this method for 11 claimants, seven other awards were recommended. Claimants Brown and Lamb, hypothetically promoted to produce manager, were given the so-called Appendix H award, including bonus. Claimants Simpson, Spencer, and Williams were given amounts which were stipulated or which could be computed from stipulated amounts. Claimant Stricklin was awarded the difference between the wages she received as meat wrapper and the wages she would have received as apprentice meat cutter, with interest. Claimant McDowell was given the difference between what she earned as a part-time clerk and what she would have earned as a full-time clerk. These “special case” awards bring the recommended award to $208,148, the Vanilla Award to $219,300 ($330,255 with interest), and the Appendix H award to $254,769. II. The Parties’ Objections. Plaintiffs have filed extensive objections to the master’s report. Their objections may be summarized as follows: (1) The reduction for probability of termination in the Appendix H award and the recommended award was improper. (2) The master should have included an inflation factor. (3) The use of average claimants’ salaries rather than actual claimants’ salaries in computing the Vanilla and Appendix H awards was inappropriate. (4) The master should have included bonuses for assistant managers and managers. (5) The failure to award a differential for nonpromotion to supervisor was erroneous. (6) Claimant McDowell should not have been limited to the higher of the differential between part-time clerk and full-time clerk and between full-time clerk and produce manager. (7) Plaintiffs urge adoption of the master’s recommendations regarding tax effects. Defendant has responded to the master’s report by letter, also presenting objections: (1) The claimants should not have been automatically promoted into the highest position (except for supervisor) they could statistically achieve. (2) The master only partially took into account the fact the male managers were soon terminated. (3) The assumption that those seeking both produce manager and assistant manager/manager promotion tracks should receive the higher of the corresponding awards is unfair. (4) The award to Claimant Strickland should have been based on the one actual vacancy and not on a hypothetical vacancy. III. The Termination Factor. The approach finally recommended by the master adjusts for the probability of termination by reducing the aggregate Vanilla Award by 40%. This figure was obtained through a complicated, statistical estimation process, and represents the master’s best estimate of the effect of the termination probability given an underlying assumption (favorable to claimants) that the termination probability begins only at the beginning of the back pay period. This reduction is based on the assumption that female managers as a group would not have performed significantly better than male managers as a group. Plaintiffs, citing Franks v. Bowman Transportation Co., 424 U.S. 747, 773 n.32, 96 S.Ct. 1251, 1268 n.32, 47 L.Ed.2d 444 (1976), argue that uncertainty about the claimants’ work performance following their hypothetical promotion must be resolved against the defendant, whose unlawful conduct created the uncertainty. They argue further that the median male tenure figures, even after adjustments by the master, are unrepresentative of the true tenure the female managerial employees would have achieved, because they include male managers who were managing at the end of the back pay period and whose true tenure is thus underestimated, and because due to discrimination the male managers are presumably as a whole less qualified than their female counterparts. Defendant argues, on the other hand, that absent evidence showing a longer tenure would have been achieved by a particular claimant or claimants, each claimant must be hypothetically terminated after the median male tenure, and that consideration must be given to the possibility that a female assistant manager would have been terminated rather than promoted to manager. The court agrees with plaintiffs that no reduction should be made for termination probabilities, but for a reason slightly different from those articulated by them. It is true that some male employees may have been terminated, voluntarily or involuntarily, due to their unsuitability as supermarket managers or assistant managers. These employees may well have been paid less than their Shop Rite managerial counterparts at the jobs they assumed after leaving Shop Rite. Indeed, they may have been paid less than the average claimant’s salary, or may have been unable to obtain employment at all. On the other hand, terminated employees may have been fully competent managers, and may have obtained managerial positions elsewhere at salaries equal to or greater than their Shop Rite salaries. In such cases, Shop Rite’s discrimination would have prevented claimants from obtaining the skills and experience which would have enabled them to obtain lucrative employment elsewhere. A pair of diagrams may be helpful: Figure 1 shows the award urged by defendant (after appropriate adjustment for the probability of failure to reach the manager level): it is assumed that the claimant would have been terminated after the median male tenure and that she would have earned a clerk’s salary at her new job, i. e., that she would have acquired no transferable skills or experience as a Shop Rite manager. Figure 2 shows the award urged by plaintiffs: it is assumed that the claimant would have remained at Shop Rite as manager or that she would have earned an equal amount elsewhere. The crosshatched portion of the figure may be regarded as the return on transferable skills or experience following a hypothetical termination. Of the two sets of assumptions underlying these models, the latter is the more reasonable. Even if large numbers of male employees were involuntarily terminated for poor performance (a fact not shown by the evidence), there is no reason to believe that those employees were forced to accept no more than a clerk’s salary at their new jobs. While some terminated employees may have had to accept lower-paying jobs due to economic conditions or to adverse inferences drawn by their new employers from the fact of their termination, others may have voluntarily left Shop Rite in order to assume higher-paying jobs elsewhere, perhaps even at the supervisory level. In the absence of evidence showing that hypothetically terminated employees would have earned a lower salary, the court must resolve doubts in favor of the claimants by awarding them a manager’s salary notwithstanding the possibility of termination. This approach is in accord with the widely-recognized Fifth Circuit rule that uncertainty in determining what an employee would have earned but for discrimination should be resolved against the employer. United States v. United States Steel Corp., 520 F.2d 1043, 1050 (5th Cir. 1975), cert. denied, 429 U.S. 817, 97 S.Ct. 61, 50 L.Ed.2d 77 (1976); Pettway v. American Cast Iron Pipe Co., 494 F.2d 211, 260-61 (5th Cir. 1974). As the court noted in United States v. United States Steel Corp., supra, “once a court has determined that a defendant’s inequitable conduct caused some damages to the class, or to a representative sample of its members, then the burden falls upon the wrongdoer to explain away or disprove the damages which each claimant’s evidence arguably supports.” 520 F.2d at 1050. The computation of back pay under these principles is within the court’s discretion, and there is no single correct formula for computing back pay. United States v. Allegheny-Ludlum Industries, Inc., 517 F.2d 826, 852 n.29 (5th Cir. 1975) , cert. denied, 425 U.S. 944, 96 S.Ct. 1684, 48 L.Ed.2d 187 (1976). IV. Inflation. New courts have squarely faced the question of whether a Title VII back pay award should include, explicitly or implicitly, an adjustment for inflation occurring between the time of the discrimination and the time of the back pay award. Such a factor has been explicitly included by one district court, see Lewis v. Philip Morris, Inc., 13 Empl.Prac.Dec. ¶ 11,350, at 6169 (E.D.Va.1976) , but that decision appears to stand alone. Other courts have denied inflationary adjustments outright, see Kinsey v. Legg Mason Wood Walker, Inc., 16 Empl. Prac.Dec. ¶ 8,168, at 4825 (D.D.C.1978) (recovery of interest and inflation factor would be double recovery), or have adjusted for inflation through an increase in the interest rate applied to the award, see EEOC v. Pacific Press Publishing Association, 482 F.Supp. 1291, 1319-20 (N.D.Cal.1979) (adjusted prime rate); Patterson v. Youngstown Sheet and Tube Co., 475 F.Supp. 344, 355 (N.D.Ind.1979) (8%). Cf. Chapman v. Pacific Telephone and Telegraph Co., 456 F.Supp. 77, 80 (N.D.Cal.1978) (inflation factor denied where salaries included cost-of-living adjustments). See also English v. Seaboard Coast Line R. R. Co., 12 Empl.Prac.Dec. ¶ 11,237, at 5730 (S.D.Ga.1975) (pretermitting question). While denial of an inflation factor in employment discrimination actions against the federal government, see, e. g., Blake v. Hoston, 22 Empl.Prac.Dec. ¶ 30,603, at 14,-235 (D.C.Cir.1980); Moysey v. Andrus, 22 Empl.Prac.Dec. ¶ 30,834, at 15,328 (D.D.C.1980), rests in part on special considerations not present in a case involving a private employer, the reasoning employed in a recent Fifth Circuit decision under the Back Pay Act, 5 U.S.C. § 5596, may be instructive. In Payne v. Panama Canal Co., 607 F.2d 155 (5th Cir. 1979), the district court ordered that back pay calculations include an inflation adjustment in accordance with the Consumer Price Indices. The court reasoned that the statutory language, specifying an award of “an amount equal” to the pay differential, must be read as authorizing an inflation factor, since a 1964 dollar was not “equal to” a 1972 dollar. Payne v. Panama Canal Co., 428 F.Supp. 997, 1001 (D.C.Z.1977). The Fifth Circuit reversed on this point, finding an inflation award inappropriate absent an express authorization for such an award in the statute or regulations. 607 F.2d at 165. Significantly, the court stated that “[t]he law does not recognize the impact on judgments of inflation occurring prior to the judgment,” citing a patent infringement case. Id. The master’s decision that Title VII claimants should not receive more favorable treatment vis a vis an inflation factor than other judgment creditors is in line with the Fifth Circuit’s reasoning in Payne. While a court in a Title VII case undoubtedly retains the power to make appropriate adjustments in the interest rate to achieve an equitable balance of factors including inflation, an explicit adjustment for inflation could create additional administrative difficulties in the computation of back pay awards, EEOC v. Pacific Press Publishing Association, supra, at 1319, and would single out Title VII claimants for treatment not accorded other creditors. As various courts and economists have recognized, an interest award includes two elements: a recovery for the “time use of money,” i. e., compensation for deprivation of the use of funds without regard to inflation, sometimes referred to as a “true” interest rate; and a factor to compensate, fully or partially, for the diminution over time in the purchasing power of the funds, i. e., an inflation factor. Expert testimony before the master establishes the “true” interest rate during the relevant period at 2-4%. Thus the 6.18% interest rate used by the master (which itself gives the claimants a slight advantage over other creditors, who would not ordinarily be entitled to compounding) includes an inflation component of 2.18 — 4.18%. From 1972 through 1978, annual percentage increases in the national Consumer Price Index for Urban Wage Earners and Clerical Workers ranged from 5.77 to 10.97%. See 4 Lab.L.Rep. ¶ 7778, at 12,931. While the inflation component of the interest award does not keep pace with these figures, the award does correspond roughly to the actual rate of return claimants could have received. The factual basis for the master’s report reveals that during the relevant period, savings yields were 4% on U.S. savings bonds, 5Vi to 5V2% on savings accounts 6 to 6V2% at credit unions. The simple fact is that the claimants, being small investors unable to invest their money for long periods of time, could not have kept pace with inflation in any event. Cf. H. K. Porter Co. v. Goodyear Tire and Rubber Co., 536 F.2d 1115, 1124 (6th Cir. 1976) (this factor is one justification for denying an inflation adjustment in patent infringement case). This being the case, the master’s proposed 6.18% figure represents a reasonable interest rate to be applied to the awards, and any further adjustment for inflation is unjustified. That the standard 6% legal rate of interest is outstripped by inflation is perhaps unfortunate; if so, however, it is a misfortune which ought to be shared equally by all creditors unless and until the main body of the law is altered to allow inflation adjustments. V. “Average” Claimant Salaries. Plaintiffs next complain of the master’s use of an “average claimant salary” for comparison to managerial salaries, rather than the actual salary earned by each claimant. They urge that this technique reduces the aggregate award by $22,023, and hence that the problem is by no means inconsequential. The figures used by the master as rough “average” claimant salaries indeed differ considerably from the averages of the salaries of the 13 claimants for whom Vanilla Awards were calculated. The master’s figures are an average of $5.71/week higher than actual average claimant salaries, rendering the weekly differential a corresponding amount too low (on the average), if average claimant salary is to be used. On the other hand, the master’s figures more nearly correspond to the median claimant’s salaries, being on the average 64<t/week lower than those figures. The purpose of the master’s approach being to compare the salary made by a typical claimant (rather than the actual salary made by each individual claimant) with that made by a typical managerial employee, and the median figures more accurately reflecting the experience of a typical employee due to the presence of a few employees earning extremely low salaries and biasing the averages downward, the use of median rather than actual figures is fair to the group of claimants as a whole. The use of median rather than actual salaries is also fair to the claimant inter sese. As noted by the master, it is reasonable to assume that pay differentials among clerks are due to factors other than chance. Absent evidence to the contrary, it is also reasonable to assume that higher-paid clerks would have become higher-paid managers. This being the case, the weekly differential should be based (as the master in fact based it) on the salary difference between a typical manager and a typical clerk. This approach, while slopping some “milk” from one bowl to another, keeps the claimants’ table dry as a whole, and gives each claimant a fair amount without rewarding those whose bowl is more than usually empty. VI. Other Aspects of the Master’s Report. The failure to award bonuses to assistant manager and manager aspirants and the failure to hypothetically promote any claimants into the supervisory ranks are urged primarily as pro forma objections at this stage. In both cases, the isolated nature of the occurrences among males renders prediction of the results which would have been achieved by claimants, absent discrimination, highly speculative. The failure to make either adjustment is therefore justifiable under the circumstances. The master limited claimant McDowell to the higher of: (1) the difference between the “average claimant’s salary” (actually the median) and the hypothesized managerial salary (i. e., the “Weekly Differential”); and (2) the difference between McDowell’s actual salary and the salary of a full-time clerk. He reasoned that it would be inequitable to award McDowell the differential between her part-time clerk’s salary and a full-time salary as produce manager. Plaintiff objects to this award, arguing that if offered a full-time position McDowell could have progressed at the same rate as other clerks. Assuming arguendo that this is so, however, the fact remains that McDowell enjoyed the benefits of a part-time position, and it would indeed be inequitable to award her more on that account than her coworkers who were shouldering a full-time load. While the tax effects of the awards are perhaps even more complex than recognized by the master, the court concurs with him and with plaintiffs that tax effects should be ignored. Defendants’ argument that claimants should not have been automatically promoted to managerial positions loses sight of the fact that the burden of showing that a claimant would not have been promoted even absent discrimination is on the employer. Watkins v. Scott Paper Co., 530 F.2d 1159, 1178 (5th Cir. 1976), cert. denied, 429 U.S. 861, 97 S.Ct. 163, 50 L.Ed.2d 139 (1976). The master’s assumption that claimants seeking both assistant manager/manager and produce manager positions would have opted for the higher-paying track is both consistent with an as- sumption of economic rationality and mandated by Claiborne v. Illinois Central R. R., 583 F.2d 143, 151 (5th Cir. 1978), cert. denied, 442 U.S. 934, 99 S.Ct. 2869, 61 L.Ed.2d 303 (1979). Defendant’s final point, regarding claimant McDowell, is foreclosed by the court’s factual finding that “a number of vacancies occurred in the Assistant [sic; Apprentice] Meatcutter position while this claimant was in the Dallas district.” Carter v. Shop-Rite Foods, Inc., 470 F.Supp. 1150, 1169 (N.D.Tex.1979). The master’s assumption of one hypothetical opening occurring midway between McDowell’s transfer into the Dallas district and the one known vacancy in January, 1974, is thus reasonable. For the foregoing reasons, the master’s report is adopted except insofar as he reduces the awards for the possibility of termination. The amounts shown in column 5 of the appendix (Vanilla Awards plus 6.18% interest, together with the master’s special-case awards) will be awarded to the respective claimants, subject to the modifications discussed in the following sections. VII. Master's Fee and Expenses. The special master, John A. Martin, has filed an application for payment of fees and expenses totaling $17,864. The parties do not object to this amount, and the court finds the master’s request to be modest given the complexity of the issues and the competence and energy brought by the master to his task. The application for payment is therefore approved. Fed.R.Civ.P. 53(a) grants the court broad discretion in determining which party or parties shall bear the costs of reference to a master. See generally 5A Moore’s Federal Practice ¶ 53.04[1], at 53-30-32 (2d ed. 1980). As items of costs, the master’s fees and expenses would ordinarily be taxed against the losing party in the litigation. See Fed.R.Civ.P. 54(d). This result may, however, be altered by the court where circumstances warrant. For the reasons which follow, the court believes that the master’s fees and expenses should be borne equally by the two sides to the litigation. Fifth Circuit precedent regarding splitting of master’s costs is itself split. In Bowen Motor Coaches, Inc. v. New York Casualty Co., 139 F.2d 332, 334 (5th Cir. 1943), the district court charged an auditor’s fee against the losing party as costs. Citing the fact that appointment of the auditor was necessary in order to permit the plaintiff to prove its case and to permit the court to make an appropriate disposition of the case, the court of appeals modified the trial court’s order to divide the expense equally between the parties. In Bros Inc. v. W. E. Grace Manufacturing Co., 320 F.2d 594, 600 (5th Cir. 1963), on the other hand, the district court divided the costs equally. The court of appeals, noting that one party had prevailed as to almost all of its contentions before the master, and that that party had neither precipitated nor prolonged the reference, held that “no satisfactory basis appears for dividing these costs,” and ordered the costs taxed against the losing party. In the present case, reference to a master was on the court’s own motion. A variety of issues were presented by both sides to the master. On some issues (e. g., automatic promotion, use of higher-paying promotion track), plaintiffs prevailed before the master and in this court, while on others (e. g., inflation, supervisory promotions) defendants were successful. On the issue of termination reductions, each side was partially successful before the master. The master’s resolution of tax effects also apparently meets with the approval of all parties. Since neither side is clearly the prevailing party before the master, and since the master’s work benefitted both sides as well as the court, it is reasonable to apportion the cost of the master’s services on an equal basis. Accordingly, while defendant is ordered to pay the master’s fees and expenses forthwith, the respective claimants’ awards will be reduced on a pro raía basis so that half of the master’s costs are borne by the group of claimants. VIII. Attorneys’ Fees and Expenses. The court’s memorandum order of May 17, 1979, provided that plaintiffs as the prevailing parties would be awarded reasonable attorneys’ fees. Plaintiffs have filed an application seeking $172,500 in attorneys'! fees, plus $15,000 for paralegal time and $3,738.23 in out-of-pocket expenses. The standards for award of attorneys’ fees in Title VII cases in this circuit are set forth in Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (5th Cir. 1974). That case lists 12 factors to be considered in determining the amount of such an award, which will be discussed seriatim. (1) The time and labor required. Plaintiffs’ counsel spent a total of 1407.95 hours of attorney time and 656.8 hours of paralegal time. Given the duration and complexity of this case, these amounts of time were reasonably necessary to adequately litigate the case. Defendant argues that this case, in which since May of 1977 it has been known that attorneys’ fees would be recoverable, has been used as a vehicle to litigate issues relevant to other Title VII class action cases in which plaintiffs’ counsel was involved, which had not reached the “Phase II” stage and hence in which an award of attorneys’ fees was uncertain. There is no evidence whatsoever that this motive prompted plaintiff’s counsel to raise any issues in this case. Even were this the motive, however, the issues raised in this case were relevant to a proper determination of the case, and are therefore the proper subject of an award of attorneys’ fees. Any incidental benefit to other pending cases is immaterial. (2) The novelty and difficulty of the questions. While some of the issues in this case were those commonly raised in Title VII class actions, a variety of novel and difficult questions were presented. Among these were procedural questions regarding class certification, questions involving the shifting burdens of proof as the litigation progressed from Phase I to Phase III, and the damage questions discussed in this opinion. (3) The skill required to perform the legal service properly. The issues discussed in (2) above required considerable skill and experience in the Title VII field for their proper presentation. \ (4) The preclusion of other employment by the attorney. Apart from that caused by the sheer volume of time consumed in this case, see (1) above, there has been no significant preclusion of other employment. (5) The customary fee. Plaintiffs’ counsel’s customary fee ranged during the pend-ency of this litigation from $75 to $125 per hour for attorney time, and $30-50 per hour for paralegal time. These customary fees are of course based on an assumption of prompt and reasonably certain payment. (6) Whether the fee is fixed or contingent. Compensation for 504 hours spent during Phase I was contingent. Compensation for 609.20 hours in Phase II was partially contingent, since a finding of liability, which substantially increased the probability of success, had been entered. Time spent in Phase III (294.75 hours plus paralegal time) was noncontingent, since the court had announced its intention to award attorneys’ fees and since some damage award was virtually certain. (7) Time limitations. As noted by plaintiffs’ counsel, time limitations were not a significant factor in this case. (8) The amount involved and the results obtained. As noted in the preceding sections of this opinion, the efforts of plaintiffs’ counsel resulted in an aggregate award of $321,325, representing 100% of the claimants’ back pay entitlements. (9) The experience, reputation, and ability of the attorneys. Plaintiffs’ counsel is one of the most able, experienced, and distinguished members of the Title VII class action bar in this district. (10) The undesirability of the case. Apart from any general undesirability of representation of plaintiffs in Title VII class actions, this case presents no special features of undesirability. (11) The professional relationship with the client. There is little prospect of an ongoing professional relationship between plaintiff and her counsel. (12) Awards in similar cases. New Title VII class actions in this district have progressed to a final back pay judgment. Testimony in other, simpler cases establishes $80 per hour as an appropriate hourly rate. Taking these factors into account, the court concludes that $120 per hour for contingent attorneys’ time (Phase I), $100 per hour for partially contingent attorneys’ time (Phase II), and $90 per hour for non-contingent attorneys’ time (Phase III), are appropriate rates. Accordingly, $147,928 (representing an average of $105.07 per hour) will be awarded for attorneys’ time. Paralegals’ time (averaging $22.84 per hour) and expenses will be awarded as prayed for. The total award for attorneys’ fees and expenses is thus $166,666. IX. Summary. As modified herein, the report of the special master is adopted. Defendant shall promptly pay the master’s fees and expenses in the amount of $17,864. Judgment will be entered for the respective claimants in the amounts shown in column 6 of the appendix to this order, together with the attorneys’ fee award set forth above. REPORT OF SPECIAL MASTER Too long ago, this Court appointed the undersigned as Special Master to determine the amount of the back pay award to each successful Claimant. So that the Court will not draw incorrect inferences from the lengthy interval, let me list the more important reasons for the delay, not necessarily in order of chronology or significance: 1) The failure to pay proper heed to Parkinson’s Law-“Work expands so as to fill the time available for its completion”; 3) Something of a fastidiousness for precise details and a compulsion toward absolute accuracy shared by all legal counsel and totally lacking in the undersigned; 2) A congenial-and necessary-accommodation by counsel and the undersigned to each of the others schedules; 4) A reluctance on the part of counsel to follow the lead of the undersigned-a reluctance soundly founded, we must add, on the correct perception that he had no real idea where in the world he was going; and 5) The nature of the beast-this is a matter better suited to the commonsense deliberation of the jury room subject to control of the Court, but one which the Title VII places exclusively in the Judge’s lap. In any event, counsel were fully cooperative in compiling the material facts in the Factual Basis for Report of Special Master (the “Basis”) which, with its three accompanying exhibits (“Ex. BA,” “Ex. CH” and “Ex. TX”), is being filed simultaneously herewith and will be referred to plenteously throughout. The following brief summary is given to explain what is the substance of the Basis and its accompanying Exhibits, hopefully in aid of the understanding of this Report. HEREIN OF THE FACTUAL BASIS The Basis begins with an overview of Shop Rite’s Dallas District during the early 1970’s-from June 6, 1971, to June 6, 1975 (the “Relevant Period” or the “Period”). It then sets out the relevant facts as to each of the groups of male employees: The Supervisory Group, the Store Managers Group, the Assistant Managers’ Group, and the Produce Managers’ Group. The information as to the Supervisors is largely contained in resumes or work histories that comprise Exhibit A of the Basis. This group is much smaller, much more diverse in employment background, and much harder to generalize about than are the other three groups. For each of the other three positions a “comparable” male group of employees was assembled for each job category. Initially the three groups consisted of all of the males who were promoted into that position during the Relevant Period. The parties then “bargained” exclusions from each group-largely because something about that individual’s work history was aberrant (a series of promotions, demotions, military service breaks, etc.) They finally wound up with the three groups as contained on Exhibits B, C and D as comparable male groups for each of the three positions. A dispute as to how to handle males with “prior comparable experience’-Claimants seeking inclusion, Shop Rite exclusion-could not be resolved by the parties, but their respective positions and their dispute is memorialized in the Exhibits and the difference in each case can be quantified. Each group is, thus, a very large sampling of Shop Rite’s male employees in the three positions for the period involved. They are comparable to the Claimants and are useful in developing the median intervals for promotion for each managerial group. For each of the three positions, median and average salaries were developed for each year. Note that it makes very little difference which salary is used to compute back-pay awards-the average or mean produced about 50$ a week more for managers’ salary, and the median about 25$ per week more for assistant managers and for produce managers, they produced the same. Similar information was set out for bonuses paid to each group on a quarterly and annual basis. As a part of the basis, it was agreed and noted that Produce Managers were on a different job track than were Assistant Managers-Managers. Only very rarely did any crossover occur. For the Claimants, individual treatment was afforded special cases and then the bulk of the claimants were considered together. For this large group-paragraphs 33,34 and 37 are crucial. Paragraph 33 sets out their claimed position, employment data with Shop Rite and their tenure at the beginning of the Relevant Period. Paragraph 34 sets out the annual wages earned by each Claimant at Shop Rite and 37 sets out every significant leave of absence during the Period. As a part of the Basis, the parties agreed that there were no significant differences between fringe benefits provided to management and those enjoyed by the Claimants-therefore, such could be disregarded. Other than some economic information about interest rates and rates of inflation, the rest of the Basis describes the three Exhibits that accompany it. Exhibit BA, contains the Bonus Accrual information-except for the actual dollar amount of bonuses paid-for each of the three managerial positions for each quarter of the Relevant Period. A study of it can determine promotions or hirings into, and exits from, the three positions. Exhibit CH is just such a study. The three complex and necessarily messy charts-one for each position-form the basis for the six summaries that make up the front part of Exhibit CH. There are two such summaries for each position-one setting out information as to “Some” managers-those promoted during the Relevant Period-and one for “Other” managers, being the incumbents at the beginning of the period. Exhibit TX contains copies of tax returns and other tax data for most of the Claimants. In addition to this Factual Basis, there are some “non-factual” or at least “non-provable” bases, and some simplifying assumptions upon which this Report also rests-which, in candor, must be disclosed. HEREIN OF THE NONFACTUAL BASES In deposing the economic expert, Dr. Swanson, Claimants’ counsel articulated the task this Court set for the undersigned as well as can be done: ... (The) objective is to award to each successful Claimant an amount of money which would compensate the Claimant for the economic loss she incurred as a result of the discrimination. In undertaking this assignment and continuing until these words are written, I have entertained grave doubts about my ability to come close to understanding what that means-much less to do it. For instance, I thought about what “the discrimination” means. It should mean the unlawful employment practices of Shop Rite that this Court has found occurred as they affected each of the Claimants-or at its broadest, it would still be limited to Shop Rite’s employment practices as such affected women generally, including the Claimants. But how do you separate out the other types of discrimination-or at least sex-based differences-that have been visited by others than Shop Rite on females? Until July 3, 1965, in most places of the United States-in particular, Texas-it was not unlawful to refuse to hire, to refuse to promote or to refuse equal pay solely on the basis of sex. And discrimination did not cease automatically and everywhere but in Shop Rite’s stores on July 3, 1965. The difficulties in isolating Shop Rite’s unlawful discrimination and its effects are reflected in the Basis in the first four Exhibits. In Exhibit A-the synopsis of Supervisors’ work history-every man (with one notable exception) had considerable pre-1965 managerial experience. In Exhibit B, C and D, the differences between the parties over inclusion into the “comparable” male group were largely over treatment of those men with “prior managerial experience.” Inclusion of males with “prior managerial experience” lowered the median tenure time for the Clerk to Assistant Manager progression by 3 months and by 1 month in each of the other two groups. As will appear below, I excluded “prior managers” across the board. None of the Claimants claimed comparable experience, and I do not believe that there existed a significant pool of prospective female employees with prior management experience. It is, of course, the fault of others than Shop Rite that its prospective female employees lacked such experience. Secondly, in the context of this case, how does one get a handle on quantifying “economic loss?” Or defining it, for that matter. Most EEO cases involving back-pay awards with which I am familiar have involved “equal work-unequal pay,” where the victim class did virtually identical-or at least comparable-work for less pay. It was not difficult there to conclude that the cents-or dollars-per hour difference should be awarded the victim class (although it seems just as reasonable to suppose that a “discrimination subsidy” boosted the favored class’ wage over a true non-discriminatory, universal wage). Although it is clear some claimants were the functional equivalent of at least assistant manager, managing is still different than clerking- and being responsible is different from having responsibility. The Claimants did not have to pay the price dn increased time commitments, mental preparation, emotional involvement, job insecurity away from a union contract, etc., that a managerial job demanded. So, if she has not made the “economic input”-although against her will and because of Shop Rite, admittedly-should she still get the full “economic loss” that the pay differential suggests? Thirdly, more importantly along that line, how would anyone ever know if-much less to what level of competence-any one of the Claimants could manage, or assistant manage or produce manage? This problem was slightly alleviated by some Claimants that aspired no higher than assistant manager. But what of those who see themselves as capable of rising to the supervising level? How can any of the Claimants prove to me they can-or Shop Rite prove to me they can’t? They can’t, neither of them. Or at least I still don’t see how they can-and I’ve thought long and hard about it. This question of proof of managerial ability is but one incidence of twin-questions that were raised several places-who has the burden of proof and what effect is a lack of proof (usually because of impossibility). Another example is the “neither road taken” problem presented by the different tracks that the Assistant Manager and the Produce Manager followed-with sometimes widely different results under our methodology. The Claimants contend that all inferences should be drawn as strongly as possible against the discriminator-citing direct quotes from numerous authorities. Shop Rite counters that it remains Claimants’ burden-despite the discrimination finding-and such burden is not met by piling inference on inference. Here are a few other non-factual or at least non-provable factors that are somehow to be taken into account-even when explicitly ignored: In computing back-pay awards, the methodology assumes that each Claimant worked the same time period after hypothetical promotion as she did as a Clerk, but promotion could have caused her to work longer for Shop Rite, quit sooner for a better job, or caused a fairly early firing. A promotion-even if short-lived-may have a significant economic benefit, not equivalent to back-pay period income in terms of increased self esteem, job mobility, etc. A firing or a demotion (which Claimants were “denied” by lack of promotion) may be an economic disaster or a spur to greater economic productivity. In short, I find it as difficult to predict the past as I do the future. And despairing of notion that what follows will be “right”- or can be-because of this host of assumptions that we have to make, I am willing to make a few more assumptions that appear reasonable to me-but more importantly sure make things simpler in an effort to achieve “rough justice” for the Claimants and for Shop Rite. SIMPLIFYING ASSUMPTIONS There were two assumptions that Claimants agreed to at the outset-amounting to concessions on their part-the first of which greatly simplified putting together the Basis and this Report: That the back-pay awards would be limited to the period-the “Relevant Period”from June 6, 1971, to June 6, 1975. No Claimant sought the position of either Regional Manager or Meat Market Supervisor. There are many other Simplifying Assumptions that I make-hopefully each explicitly done-throughout this Report. There are these few-however-that are of such crucial importance that they should be stated at the outset: The Basic Assumption-The reliability of a crude form of Probability Theory that I find much easier to draw upon than to articulate. The notion is that our group of Claimants is large enough and our group of comparable males is certainly large enough (almost universal) and representative enough that we can get a fairly good picture of the experience of a typical or average Claimant, and the comparable male. Maybe a fuzzy picture, lacking in detail, but a good shot at reality anyway. The Corollary Assumption-That this typical picture will not look like any real Claimant and cannot be used mechanically for any one Claimant, but should work quite well for the group of Claimants as a whole. For example, five Claimants lay claim to a Supervisor’s position-it is one thing to say that the odds are pretty good that one, two, maybe three Claimants could have reached that level than it is to say that Patricia Carter or Joyce Debord or both or neither would have been the one that did. The Damage Assumption-1 do not believe that there is any necessary relationship between what damages Shop Rite should-in fairness-be required to pay or what-in fairness-any one Claimant is entitled to receive. In other words, it makes as much sense-and is no harder to do (or no easier) to first determine how much Shop Rite should pay and then apportion it among the Claimants as it is to determine what each Claimant should get and then aggregate those amounts to determine what Shop Rite should pay. This has to be true with any exemplary damages that form a part of a Title VII recovery. If Discriminator is ordered to pay $100 exemplary, where is it written that each of the hypothetical Claimants must share on a pro rata basis rather than on an amount proportionate to their actual damages (i. e., total back pay award?) or on some other, perhaps more equitable, basis. And, if as has been held there is room for the equitable power of the Court to fashion Title VII relief, where is it written that those equitable powers are similarly circumscribed. The Consequent Assumption-As a consequence of the three assumptions, I am further willing to assume that as long as the overall recovery is fair and the Claimants are treated fairly vis a vis each other, we don’t have to agonize too long, or compute too finely. The Arbitrary Assumption-Given the above assumptions and under the above constraints, I believe that a neutral other can play some type of God by making gross, even arbitrary assumptions as to the workings of that hypothetical world that is the subject matter of this lawsuit, just as long as he is as happy to slash at Shop Rite as he is to chomp on the Claimants. Which I am. The Unscrambling Assumption-'We proceed largely through discussion in the text and calculations or proof in the Appendices. Our whole structure is built upon the concept of a “Vanilla Back-Pay Award” which is developed in Appendix A. It shows for most of the Claimants-other than some special cases-what she would receive based on applying the median male tenures in her claimed position-other than Supervisor-to her time of employment at Shop Rite to determine her time in the particular managerial position and assuming automatic promotion after such time for those aspiring Managers. Next a differential between Clerk’s wages and the position’s salary is multiplied by the time in the position to get the actual back-pay. It is plain vanilla-without a time-use or interest factor or a tax factor and with no other adjustments or reductions. If the Court does not buy in whole or in part our recommendations then it will be easy to correct for the differences by using the figures in Appendix A and adding-or eliminating as the case may be-the nuts and the syrups. USE OF “AVERAGE CLAIMANT SALARY” CONCEPT Another simplifying assumption I make at the outset in compiling Appendix A is that of the “Average Claimant Salary.” Paragraph 34 of the Basis sets out the actual weekly average salary for each Claimant for each year-at least insofar as such information could be reconstructed. Therefore, they are reasonably-but not absolutely-accurate. For most of the Claimants who worked an entire year without lengthy illness or other leave time, the figures approach complete accuracy quite closely. For Claimants who were ill, otherwise on leave, or who voluntarily or otherwise worked part-time, I am much less confident in the accuracy of any figures taken individually. At least three things can be done with the Paragraph 34 figures. One, they can be used as they are presented to determine-by subtraction-the difference that Claimant would have earned in the particular manager’s position. This penalizes the higher paid Claimants-who presumptively are the more dedicated, ambitious etc.-in relation to their lower paid counterparts. This unfairness is particularly noticeable when considered against the distribution of Paragraph 34’s figures. See Appendix B for a study of claimants’ salaries. Most of the Claimants most of the time fall in a narrow range at the top and the rest fall-not just below-but far below this range. The second possibility is to develop an average or a minimum-for example, the bottom of the range referred to above or the mean-and apply it to bring those below it up to that level, but to use the actual figures for the higher paid Claimants. This is fairer than the first possibility because it does not reward accidental or habitual under-achievement. But it still penalizes by a loss of earnings the higher-paid Claimants. And it involves still a lot-though less-special calculations. The fairer thing seemed to be to develop an “average” figure for each year as the salary for the position of Clerk and apply it to all Claimants. One, it’s easier to calculate-much easier; particularly since I used a whole dollar figure for our “average.” Secondly, it does not penalize the higher-paid Claimant. Thirdly it comports with what I perceive to be the relative rather than absolute, accuracy of the Paragraph 34 figures. And it does not provide a windfall to the higher-paid Claimants-although at first blush, it seems to. It must be remembered that the other figure involved in the subtraction process referred to above is the median for that managerial position, and there is reason to believe that the higher-paid Claimants would perform above the median (given the assumption that the Claimant group as a whole would perform identical to the median). Lastly, it’s not too much hair off Shop Rite’s hide-and it is actually much better for it than using the actual wages for all Claimants. For example, a quick calculation showed that Ruth Brammer-the highest-paid Claimant-gets about $1,200 more (less than $8.00 a week) using the “average” than she would have had her actual figures been used. It reduces what 5 Claimants-three of them substantially-would get if their own actual figures were used; it makes less than $1.00 per week difference for 2 more and about $2.50 per week for the rest. As Joe Stephens observed about milk consumption for his 11 kids: “We spill more than that for breakfast.” The figures I developed for the “average” weekly Clerk’s salary-by extrapolation from Paragraph 34 of the Basis-for each of the years are as follows: 1971 $125.00 1972 $130.00 1973 $145.00 1974 $160.00 1975 $190.00 In each year there were two actual Claimants making the amount “developed” except 1974 when there was only one; there were at least twice as many below the “developed” amount in each year, again except 1974 when there were 6 above and 9 below (which indicates the fairness to Shop Rite of this composite approach). And it is in rounded numbers which are easier to use in computations. Finally, it compares well-surprisingly well assuming a little “filter-down” time-with the increases in the median weekly Store Manager’s salary set out in paragraph 12: Annual Increase 1972 1973 1974 1975 Clerk’s "Average" 4t 11.5% 10.34'% 18.75% Manager's Median 8.43% 8.24% 11.66% 6.12% Increase Over 1971 Base Clerk's "Average* +4% + 16% + 28% ♦ 52% Managers' Median + 8.43% + 16.91% + 30.54% + 38.53% Given the abnormality of 1975 when Manager’s revolved and clerks that wanted to could undoubtedly work overtime while their cohorts sought permanent employment elsewhere, this comparison reinforces the accuracy of the developed average figure for weekly clerk’s salary. Finally, the reasonableness of the concept’s use seems established by Appendix B. THE WEEKLY DIFFERENTIAL The developed average weekly Clerk’s salary can now be subtracted from the appropriate weekly salary to get the weekly differential for each position for easy year. For Manager, the “Median/Weekly” figures in Paragraph 12 were used-which are without bonus. For Assistant Manager, the “Median/Weekly” figure in Paragraph 16-again without bonus-were used. For Produce Manager a weekly bonus amount was determined by dividing the “Median/Quarterly” figure in Paragraph 21 by 13 (weeks per quarter) and this figure was subtracted from the “Median/Weekly” figure in Paragraph 20. The weekly differentials are, then, when rounded to the nearest whole dollar, as follows: CLERK-STORE MANAGER DIFFERENTIAL Weekly Year Clerk's Weekly Manager's Weekly Differential 1971 $125 206.84 $62 1972 130 223.40 93 1973 145 241.81 97 1974 160 270.00 110 1975 190 286.54 96 CLERK-ASSISTANT STORE MANAGER DIFFERENTIAL Year Clerk's Weekly Assistant Manager's Weekly Weekly Differential 1971 125 178.47 53 1972 130 185.84 55 1973 145 193.21 48 1974 160 207.87 47 1975 190 225.00 35 CLERKS-PRODUCE MANAGERS DIFFERENTIAL Year Clerk's Weekly Produce Weekly Manager's Weekly Differential 1971 125 157.36 32 1972 130 165.65 36 1973 145 178.49 33 1974 160 197.96 38 1975 190 190.92 1 Despite the apparent absurdity of the last figure entered above, I adopt all of the above as the most reasonable figures available to determine differentials in weekly salary-without bonuses-among the various positions. They will be applied to develop the Vanilla Individual Back-Pay Award (the “Vanilla Award”) for each Claimant. FUNCTION OF THE VANILLA INDIVIDUAL BACK-PAY AWARD As stated, Appendix A computes a back-pay award for each Claimant (other than McDowell, Spencer, Simpson, Stricklin, and Williams) without taking into account any bonuses for each position she claimed and disregarding any supervisory position, i. e., anything above Store Manager. Remember that District Manager and Group/Produce Supervisor were the two supervisory positions sought by the Claimants. The methodology used to compute the final figures for each Claimant is more fully described in the introduction to Appendix A; broadly stated it is a computation for each Claimant utilizing the concept of a median time for male employees for progression from position to position and the Claimant’s actual employment history with Shop Rite, beginning frequently before the Relevant Period starts, but ending simultaneously therewith. It is the basis for the ultimate award to each of these Claimants-but the rest of this Report does make adjustments to those figures. Most of the Claimants were treated as incumbents in one of the three managerial positions at the beginning of the Period. This stems directly from findings the Court has already made as to the turnover during the Relevant Period in Shop Rite’s managers and which is easily observable from Exhibits TX and CH. I have made the assumption that the turnover in all three positions was similar from passage of Title VII to the beginning of the Relevant Period; in fact, turnover could have been much less and still have had the same results for our purposes: there would still have been plenty of openings for a Claimant to have filled shortly before, at or after she had served as long in position as the median male. Those Claimants who served through the median male tenure for promotion, became, under the theory, eligible for promotion at various times beginning as early as 1966. Such turn-over during the Relevant Period leads likewise to the reasonable conclusion that at the time within such Period a Claimant served through the male-tenure time, there would have been at least one, probably many openings available then or shortly thereafter (or before since it “averages” out). Consideration should be given to the stated presumption that between 42% and 50% of the Shop Rite work force was female and to the direction that such ratio guide the compiling of this Report. Shop Rite had 25 stores for the first couple of quarters, but 5 of them closed by or were closing in early 1972 when # 340 became part of the Dallas District. Thus, 20 stores were in operation the bulk of the Period. Seven Claimants, under our method, were deemed Managers for the entire period. One of the other eleven Claimants managed at the beginning of the Relevant Period, not at the end-and the other three au contraire. The minimum number of female Managers at any one time during the Period was 8, and for over half of the period, 10. This bumps pretty hard against the top of the Court’s ratios. There is no similar problem with Assistant Manager or Produce Manager-and the bottom of the range is not approached if all those positions are combined. SUPERVISORY PROGRESSION Only if the Court is willing to carry the Theory of Proportional Sexual Representation to the highest-i. e., the supervisory-level, is there anything to support taking supervisory pay or position into account in fashioning the back pay awards. While I am unwilling ultimately to make this assumption, I am willing to entertain it temporarily to see where it might lead. The Court found that at least 13 openings occurred at all Supervisory levels except Regional Supervisor during the period it had under discussion which extended to 1977. From the resumes of supervisory personnel that make up Exhibit A it can be determined that at least the following vacancies occurred in the following supervisory positions: District Manager. Sam Allen who was promoted to District Manager in August of 1970 was “terminated” (retired?) in April of 1972, and was replaced by James Spears, who served until the end of the year and assumed another Supervisory position. Buddy McCaghren, who replaced Spears, served apparently until January 1, 1975 when he was replaced by Bobby Smith. There were, thus, three openings in this position. District Produce Supervisor. Doyle Stewart who entered this position in November of 1970, was terminated in April of 1972. Apparently this position was filled three months later by Jack Gates, who served until the end of the period. There was, thus, only one opening in this position. District Grocery Supervisor. In September of 1971 McCaghren filled the position of grocery supervisor until promotion to district manager at the end of 1972. His successor was Allen Seale, who served the first 10 months of 1973 until he was terminated and replaced by Bobby Smith, who served until January of 1974 when he became Dallas District Manager. There were four openings for this position in the Relevant Period assuming that the position was not filled after Smith occupied it. Thus, with women occupying about 42% of the work force then, under the Theory they would occupy about 42% of these positions-or fill a